Prediction allows knowledge and experience to guide action and is critical for a range of sensory, motor, and cognitive functions. Failure to generate accurate predictions could contribute to neurological disorders such as autism and schizophrenia. This proposal takes advantage of a simple model system in which it is possible to dissect the cellular and circuit mechanisms for generating predictions and to understand their functional roles. Weakly electric fish possess a specialized organ in their tail that generates an electrical field and specialized electroreceptors that are sensitive to small changes in the strength of the field. Detecting changes in the field induced by nearby objects allows the fish to navigate and find prey in darkness. However, because the electric organ is in the tail and electroreceptors are located on the head and trunk, the fish's own movements also alter patterns of electrical inputs. Hence, the challenge for the electrosensory system is to distinguish between behaviorally relevant patterns of input due to external events from patterns that are self- generated. Though particularly clear and accessible to study in electrosensory systems, this same problem faces sensory systems in any animal that moves. For over a century scientists and philosophers have puzzled over how we perceive a stable visual world despite the fact that visual input changes dramatically several times per second due to rapid movements of the eyes. One possible answer is that the brain generates predictions about changes in visual input that will result from our own movements and filters out these predictions from the actual sensory input. Previous studies have shown that just such a process occurs in a region of the brain of electric fish that closely resembles the cerebellum. Such predictions are formed via changes in the strength of connections between neurons, a process known as synaptic plasticity. Virtually identical synaptic plasticity mechanisms exist in the mammalian cerebral cortex and cerebellum and likely underlie learning and memory. This proposal uses neural recordings and computational modeling to provide insith into two general issues. The first will test the hypothesis that cerebellar granule cells (the most numerous neurons in the vertebrate brain) provide critical 'raw material' needed for forming associations (via synaptic plasticity) between aspects of the fish's movements and the resulting predictable patterns of incoming electrosensory input. Second, the proposed studies will establish roles for sensory and motor signals in the generation of predictions in the context of sensory filtering. Detailed links between synaptic plasticity, neural circuitry, and sensory function and will provide key insights into both cerebellar function and the neural mechanisms for predicting sensory events.